Update to nightly-2023-11-25

[bjorn3]

No description provided.

[bjorn3]

I'm not sure if the changes in this file are correct.

[bjorn3]

Locally I'm getting a test failure which looks like UB or a miscompilation somewhere. Basically for the arg test the output is empty, but if I change the println!() to eprintln!() it does show up on stderr. Furthermore if I eprintln!("{:?}", args), allcaps is shown as true, but if I remove the if args.allcaps { msg = msg.to_uppercase(); }, it suddently changes to false.

[willcrichton]

Thanks for the PR. (Just curious, are you using rustc_plugin for something?)

I will handle bumping the nightly (see #16). I decided to delete the entire places_conflict module because the new rustc_borrowck::consumers module now provides the same functionality publicly: https://doc.rust-lang.org/nightly/nightly-rustc/rustc_borrowck/consumers/fn.places_conflict.html

[bjorn3]

I was looking into updating flowistry to the latest rustc version. I have been working on spme signficant refactors to the driver interface and wanted to make sure I wouldn't break flowistry.

[willcrichton]

Sorry I let this drop, but I'll just add -- @bjorn3 don't worry about breaking Flowistry, I am very used to it breaking every time the compiler updates 🙃 a changelog is more than enough!

[bjorn3]

In this case it is a non-trivial change of the rustc_driver interface that I fear will break flowistry in a way that could potentially require a rustc change to fix it again. To be precise I want to get rid of the rustc_driver query interface.

[willcrichton]

Got it. If you want to test anything, I just updated the dev branch of Flowistry to nightly-2024-01-06.